Operator order run using Multiply affects validation accuracy?

thinhakapetethinhakapete Member Posts: 1 Learner I
edited November 26 in Training Q&A Forum

Hi guys,

I am trying to learn Validation and Performance Measurement in Foundations Course.
One observation I have found that my Performance accuracy of my process is slightly different from the Solution file. What I also find out is the order of operator run will make 2 validation accuracy change despite I did use Multiply operator (which is described 2 datas won't affect each other).

What is the reason for this? Is it a bug or something else?

 

Compare.jpgOperator Order in Multiply affects Validation accuracy?

 

Best Answer

  • mschmitzmschmitz Posts: 1,824  RM Data Scientist
    Solution Accepted

    Hi,

     

    it's fairly simple. The measurement of performance depends on the random seed. Thus you also have this (large) standard deviation. Just fix the local random seed and you will get the same results in every execution.

     

    Best,

    Martin

    - Head of Data Science Services at RapidMiner -
    Dortmund, Germany

Answers

  • sgenzersgenzer Administrator, Moderator, Employee, RapidMiner Certified Analyst, Community Manager Posts: 1,867  Community Manager

    tagging @Knut-RM

     

     

Sign In or Register to comment.